Clin Res Cardiol (2023). https://doi.org/10.1007/s00392-023-02180-w

The association of NT-proBNP levels with left ventricular enddiastolic pressure stratified by ejection fraction in patients with heart failure

M. Lehmann1, M. Mendez-Bräutigam1, J. Urban1, S. Schulze2, O. Riedel1, K. Nobis-Joachim1, C. Friedrich3, K. Huenges3, A. Haneya3, J. Strotmann1

1I. Medizinische Klinik, Städtisches Krankenhaus Kiel GmbH, Kiel; 2Zentrallabor, Städtisches Krankehaus Kiel GmbH, Kiel; 3Klinik für Herz- und Gefäßchirurgie, Universitätsklinikum Schleswig-Holstein, Kiel;

Background: Heart Failure (HF) is an increasing clinical problem and its diagnosis is a challenge for many healthcare systems especially in patients with preserved ejection fraction (EF). Therefore, different diagnostic algorithms are proposed to make the diagnostic pathway easy and robust. In all those algorithms the biomarker NT-proBNP plays a crucial role as a surrogate marker for left ventricular enddiastolic pressure (LVEDP). Methods: We performed an all-comers registry study  including over 1000 patients undergoing left heart catheterization for different clinical indications. All patients had an invasive measurement of LVEDP and assessment for NT-proBNP at the same time plus a comprehensive echocardiography within 24 hours of the heart catheter procedure. Patients with acute coronary syndromes including NSTEMI and STEMI were excluded from the registry and patients with NYHA Stage 0 were not included in the analysis. All patients gave a written informed consent in participating in the registry. Results: A total number of 991 patients were included in the analysis. All patients  had signs of heart failure stage NYHA I to IV and 14.3% had a history/diagnosis of atrial fibrillation.  A total number of 182 (18.4%) had an EF < 50% with a median of 37% (26;43) and 809 patients had an EF ≥ 50% with a median of 62% (58;66). The median NT-proBNP level in patients with EF < 50% was 1635 pg/ml (651/3013) compared to 171 pg/ml (55;573) in patients with EF ≥ 50% (p< 0.001). Median LVEDP was 16.5 mmHg (12;23) in patients with EF <  50% and 15 mmHG (11;20) in patients presenting with EF ≥ 50% respectively (p< 0.001). A total of 60.4% of patients with an EF < 50%  had an LVEDP ≥ 15 mmHg compared to 50.2% of patients in the EF ≥ 50% group (p=0.01). For both patient groups there was a significant difference in NT-proBNP levels according to the LVEDP being < or ≥ 15mmHg (see fig. 1). After adjustment for sex and age, but including patients with atrial fibrillation, the best cut off for NT-proBNP predicting an LVEDP ≥ 15 mmHg in patients with EF < 50% was > 300pg/ml (OR 5.45 (1.89-15.73) p≤0.002) and for patients with an EF ≥ 50% it was 125 pg/ml (OR 2.08 (1.53-2.84); p≤0.001). After exclusion of all patients with atrial fibrillation in the group with an EF < 50% the odds ratio for predicting an LVEDP ≥ 15 mmHg was still best at an NT-proBNP level of > 300 pg/ml  (8.61; p≤0.001). Looking at the patient group with an EF  ≥ 50% after exclusion of all patients with atrial fibrillation it could be documented  that the odds ratio was highest (OR 2.22; p≤0.001) for an NT-proBNP level of ≥ 125 pg/ml. Conclusion: There were significant difference in NT-proBNP levels between heart failure patients with EF < 50%  and those with EF ≥ 50%. When subdividing within both groups according to the LVEDP level of < or ≥15 mmHg there was still a significant difference in median NT-proBNP levels, but in both patient groups with an EF < or ≥ 50%  there was a substantial overlap of the documented NT-proBNP ranges. Further analyses are needed to evaluate and interpretate  the role of NT-proBNP to predict increased LVEDP levels in the individual patient in a clinical setting.

https://dgk.org/kongress_programme/jt2023/aP1735.html